Big data, the term which became popular around 2011, has got bigger: big data, streaming data, open data, high frequency data, and alternative data are just some of the creatures inhabiting this new world. Very large quant funds will have in-house data strategy teams who will often contact companies directly and then evaluate if their data sets are of interest for research purposes and for back testing and so on. But the majority of smaller shops are likely to be in the process of building their capabilities and do not have these types of resources.
Rado Lipus, Neudata
Lipuš said: “It takes a lot of resources to back test some of these data sets and a lot of these data sets are quite dirty and they need a lot work and need to be tagged, they need to have tickers, and structure. So there are only a few well-funded data-centric funds that have those in-house capabilities. For most of them, they are realising the importance and they are building teams and in-house capabilities.”
The flipside is that many companies, which are not established data providers, are looking to monetise their data and don’t yet have a network of clients. “We talk to hundreds of data providers, and these are not necessarily well-established, often just starting out maybe with zero clients, or maybe with one or two clients, sort of early days”.
“It could be a data set which is a new one for an existing provider, and the nature of the data set is quite different or the use case is quite different. Often these companies struggle. They don’t know how to go about it; how to reach the right audience, but also where the value of their data is. Often these data providers know very little about the asset management or quantitative investing world. So we fill in this knowledge gap between both sides.”
Neudata does this by networking and looking for niche providers. Traditionally hedge funds go to conferences to connect with data providers for instance. However, these are typically more established data providers, which already have a strategy and already have sales people and will have done some PR, says Lipuš.
“We are less interested in those providers that are already at conferences,” he said. “We are interested in newer and less known entrants – it could be a Fintech company or it could be a listed company who is thinking to monetise a lot of their data.”
For example, Lipuš is talking to a large family office whose combined shipping, air freight and logistics businesses represent 10% of daily trade data across the globe. “This family office decided to monetise the data and so they put a team together and they combined all the data from all the different entities into one big data centre. Then they found a way to make it compliant and make it legal; sort of obfuscate the data but also aggregate it so you can’t tie it back to one specific company.
The family office in question made the data available in May this year and signed up four hedge fund clients, but once their network was exhausted they struggled to sign up any more, said Lipuš.
“Then they actually went out and tried to find people who can help them and so they found us. We see this happening quite a lot. Initially you can go to various funds which are very public about it and they have a sort of high profile. But it’s hard to reach a wider audience.
“A lot of people know what we do now and it’s just word of mouth. We work with various other providers who do a lot of data work for companies and often know of people who are trying to monetise the data, say in the App world. Many Apps are trying to monetise the data; Foursquare is a great example. They are very public about it. For them the monetisation of their data is very much selling the data to the investment world.”
In terms of interesting data sets which his firm is scoping out, Lipuš said the form that certain transactional or receipt data takes in the US, where receipts get picked up electronically at various places, is being prepared in Europe and the UK by some large supermarket chains.
“We will see more and more of that happening; so it’s not that new for people in this space in the US, but probably for Europe it is something which will be available for the first time. So we are quite excited about those data sets. We have asked a few clients and there is obviously a lot of interest in that.”
Lipuš said another company he deals with in New York started out two years ago with no hedge fund clients; they were mainly serving corporate clients. “They were initially approached by one large quant fund in the US which paid them $5K-$10K per month for the data. Today they have 30 clients subscribed to the data set and it’s become their main revenue stream – no sales people, no advertising, it’s all word of mouth.”
One commonly held notion is that when too many people find out about a certain data set it loses its alpha generating cachet. The reality is more nuanced says Lipuš.
“We know of one data set with 50 subscribers. However, the data set is used completely differently by everybody, from an intra-day to a six month time horizon strategy.
“Also a lot of these data sets are not like price data from Bloomberg. They have a lot of different fields and quite complex structures. You need to slice and dice that data and prepare it before you can actually have a time series which you can use for your analysis.”
As mentioned above, Neudata is hiring for sales executives and data researchers, and would like to engage with any firms looking to monetise their data.
Data scientists out there may be interested to hear that Newsweek and IBTimes UK are to host an Artificial Intelligence and Data Science in Capital Markets event, taking place between 1 and 2 March 2017 at the Barbican in the City of London.
